HISTORY OF RADIOCARBON DATING W.F. LIBBY DEPARTMENT OF CHEMISTRY AND INSTITUTE OF GEOPHYSICS, UNIVERSITY OF CALIFORNIA, LOS ANGELES, CALIF. , UNITED STATES OF AMERICA Abstract HISTORY OF RADIOCARBON DATING. The development is traced of radiocarbon dating from its birth in curiosity regarding the.effects of cosmic radiation on Earth. Discussed in historical perspective are: the significance of the initial measurements in determining the course of developments; the advent of the low- level counting technique; attempts to avoid low-level counting by the use of isotopic enrichment; the gradual appearance of the environmental effect due to the combustion of fossil fuel (Suess effect); recognition of the atmosphere ocean barrier for carbon dioxide exchange; detailed understanding of the mixing mechanism from the study of fallout radiocarbon; determination of the new half-life; indexing and the assimilation problem foi the massive accumulation of dates; and the proliferation of measure­ ment techniques and the impact of archaeological insight on the validity of radiocarbon dates. INTRODUCTION The neutron-induced transmutation of atmospheric nitrogen to radiocarbon, which is the basis of radiocarbon dating, was discovered at the Lawrence Radiation Laboratory in Berkeley in the early thirties. Kurie [la] , followed by Bonner and Brubaker [lb] , and Burcham and Goldhaber [lc] , found that the irradiation of air in a cloud chamber with neutrons caused proton recoil tracks which were shown to be due to the nitrogen in the air, and in particular to the abundant isotope of nitrogen of mass 14. The neutrons producing the tracks appeared to be of thermal, or of near thermal energy, and the energy of the proton therefore gave the mass of the 14C produced; from this it was concluded that the beta decay of 14C to reform 14N from l4C should release 170 kV energy (2. 7 X 10-7 erg). So the work of Kurie, Bonner, Brubaker, Burcham, and Goldhaber laid the foundation in the reaction 1 14 n + 14N = H+ C (1) The thermal-neutron absorption cross-sections for the elements were measured systematically by the Rome group, under Fermi, and it was found that the element nitrogen had an effective cross-section of about 1.7 b (one barn is 10-24 cm2) which is large, compared to most materials. This indicated that the (n, p) reaction (1) is unusually probable and that, in fact, thermal neutrons in air would be expected to be converted essentially quantitatively into carbon-14 by this reaction. At this early time, in 1936, the Table of Isotopes showed that there was a blank at the position mass 14 in the element carbon, and indicated therefore that there might be a radioactive isotope of this mass. There­ fore, following on the reporting of the (n, p) reaction with nitrogen, it seemed reasonable to try to produce radiocarbon by irradiating nitroge- 3 4 LIBBY nous materials with neutrons. This was attempted by the author's group at the University of California at Berkeley by placing 100 kg or so of ammonium nitrate near the target of the newly completed cyclotron in the Lawrence Radiation Laboratory. The idea was that the neutrons would produce radiocarbon which would be trapped in the ammonium nitrate crystals as carbon monoxide, or carbon dioxide, and which would be released on dissolving in water. In the early research of Yost and collaborators [2] at the California Institute of Technology it was shown that carbon produced by a similar reaction behaved chemi­ cally, as one would expect, and formed CO and C02. Therefore, it was supposed that any radiocarbon produced by the neutrons could be collected and measured for radioactivity. This was undertaken by Samuel Ruben for his doctoral thesis under the direction of the author in the Chemistry Department at the University of California at Berkeley. However, it was mistakenly supposed that the number of atoms needed would correspond to something like the case of sulphur-35, with which experiments had been made. Sulphur-35 has an average life of four months and it was thought that, since the 170 kV decay energy expected for radiocarbon was close to that observed for sulphur-35, the lifetime would be similar. Bombardments were therefore planned on the assumption that this would be the case. Actually, of course, the life­ time is 8300 yr on average, so Ruben and the author made only 1/25 000 of what was needed; therefore there was failure to detect any radiocarbon produced in this first experiment. It is known today that the procedures used were quite adequate and radiocarbon would have been found if it had been irradiated more intensely with neutrons; but the attempt was doomed to failure because of ignorance of the long lifetime. It is fair to say that, even today, the reason for the long lifetime is unknown, and in those early days there was discussion of very rough relationships between transition energy and lifetimes in analogy to the case for alpha radioactivity where tight relationships hold. Of course, it is most fortunate for radiocarbon dating that the lifetime is so very unexpectedly long. Ruben went on to take his degree with the author on other subjects in physical radiochemistry and, after having finished, joined forces with Martin Kamen, who had come to the Lawrence Radiation Laboratory from Chicago a couple of years before to have another try at radiocarbon. This time they succeeded [3] . By bombarding graphite with a strong deuteron beam from the cyclotron, the (d, p) reaction on the 13C present in the natural mixture gave enough radiocarbon for them to detect. On the basis of this they gave the tentative value of 25 000 yr as the lifetime. This was in 1940. World War II came, Ruben died, and after the war Kamen went on to biochemical work. However, before his death, Ruben published a most distinguished series of papers on the use of radiocarbon in the study of photosynthesis. The next step in the history of radiocarbon dating was the discovery by Korff that neutrons are produced in the atmosphere by cosmic rays. A counter had been developed at Berkeley by Korff and the author's group which was capable of detecting neutrons [4] ; they found, on flying this counter on a balloon, that its count rate increased with altitude to a maximum at some 50 000 ft, after which it fell off again (Fig. 1). It was a reasonable assumption at the time that the neutron was radio­ active with respect to the formation of hydrogen, but its lifetime was SM-87/40 5 unknown. (Subsequently it was shown that the mean life is 18 min. ) It was clear, therefore, that there would be little chance that the neu­ trons found would be cosmic-ray primaries, since they would not have been able to survive the long times in flight which cosmic-ray primaries undoubtedly require. Further, the fall-off at the top of the atmosphere was conclusive evidence that they were not primaries for, had they been, there would have been little, if any, fall-off; it appeared that the fall-off must be due to their having escaped the Earth. In his first reference to this work Korff [5a] pointed out how the (n, p) reaction on nitrogen would undoubtedly make carbon-14; from the data of Korff and Hammermesh [5b] it was possible to estimate that, on average, one or two atoms of carbon-14 would be produced in this way each second for each cm2 of the Earth's surface. 600 400 200 100 "~«—» — 60 ^ 40 •v. ft 20 ^\. 10 6 " 'V, 4 \. 2 | \ | ^ )| 1 01 20 | 3 0 4 0 5 0 6 0 I00C 00 500C) 0 3 0000 FE ET FEE T FEET 70000 FEET PRESSURE IN CM Ho FIG.l. Neutron density in free atmosphere versus altitude at Princeton, N.J. (8 Jan. 1949) (Phys. Rev. J74 (1948) 504; J6 (1949) 1267; T7 (1950) 728; Bull. Am. Phys. Soc. 23 2 (1948) 21) After World War II the story was taken up again when, in a general study of the effects of cosmic rays on the Earth, the author's group at Chicago decided to concentrate on carbon- 14 and tritium made by the cosmic-ray secondary neutrons in the atmosphere. In an earlier research at Berkeley [6] Cornog and the author had shown that fast neutrons on nitrogen make tritium (radioactive hydrogen of mass 3 and mean life 18 yr) and carbon-12, somewhat in analogy to reaction (1) by which slow neutrons on nitrogen make ordinary hydrogen of mass 1 and carbon-14. The plans for this research were outlined at that time [7] . HISTORY OF RADIOCARBON DATING The real beginning of the history of radiocarbon dating as such was in the realization that the cosmic-ray production of radiocarbon in the 6 LIBBY high atmosphere leads to a continuous labelling of the biosphere and living matter, which is terminated at death. It is difficult to know exactly when this idea was born, but it was very soon after the author's plan was fixed to detect the effects of cosmic rays on the Earth's atmosphere. It was necessary to know how a test could be made for the radiocarbon that must be present on Earth, according to Korff's measure­ ments; in considering where to look for this natural radiocarbon, and how to detect it, it was realized that the supply of radiocarbon from the well-mixed system, consisting of the carbon dioxide in the air, and the dissolved salts in the ocean and in the biosphere, would be cut off to any living being at the instant of death.